EP2502296B1 - Assembly for a fuel cell and method for producing same - Google Patents

Description

The invention relates to an arrangement for a fuel cell, with an electrode and an electrolyte, and to a method for producing the arrangement.

In the production of high-temperature fuel cells, a substrate is usually used, on which an electrolyte and the two electrodes (cathode and anode) are applied. For example, first the anode, then the electrolyte and finally the cathode are applied to the substrate. These components of the fuel cell, which are applied in layers, are electrochemically active cell layers and are also referred to as cathode-electrolyte-anode units (KEA units), as described, for example, in DE 103 43 652 A1 is known. The substrate acts as a mechanical support for the KEA unit and is, for example, ceramic or metallic.
In DE 103 43 652 A1 a metallic substrate is provided, for example a porous body consisting of sintered or pressed metal particles. Metallic substrates have the advantage that they enable a good thermal adaptation to a so-called interconnector and a technically simple electrical contact with this interconnector. The interconnector - also known as a bipolar plate or current collector - is arranged between two fuel cells and connects the individual fuel cells electrically in series. The interconnectors also mechanically support the fuel cells and separate and guide the reaction gases on the anode and cathode sides.

The electrolyte is arranged between the anode and the cathode. The electrolyte has to meet several requirements. It must conduct oxygen ions and at the same time be insulating for electrons. In addition, the electrolyte must be gas-tight. In addition, an undesirable chemical reaction between the electrolyte and an adjacent electrode must be avoided. In order to meet these requirements, in DE 10 2007 015 358 A1 a multilayer structure of an electrolyte from at least three layers is provided.

Out DE 196 26 342 A1 A method for producing a thin electrolyte layer on a porous electrode is known, in which a suspension comprising electrode material is first poured onto an electrode and dried (intermediate electrode layer), the size of the solids content of this suspension being selected such that after drying, the intermediate electrode layer is one has an average pore size that is smaller than the average pore size of the electrode layer. A further suspension comprising electrolyte material is then poured onto this intermediate electrode layer and dried.

The invention has for its object to provide an arrangement that simplifies the construction of a fuel cell. The invention is also based on the object of providing a method for producing such an arrangement.

This object is achieved by an arrangement with the combination of features of claim 1 and by a method for producing the arrangement with the combination of features of claim 13. According to the invention, an adaptation layer is provided in the arrangement between an electrode and the electrolyte. This adaptation layer brings about a good connection or adaptation of the electrolyte to the electrode. Furthermore, it supports a flat structure of the arrangement or the fuel cell if a metallic porous carrier substrate is provided for the electrodes and the electrolyte.

On the one hand, metallic porous carrier substrates are mechanically more stable in comparison to ceramic carrier substrates and can be provided with a particularly small substrate thickness. On the other hand, the gas-tight electrolyte should be made as thin as possible. This requires the smallest possible roughness on the electrode surface associated with the electrolyte (eg anode surface). Accordingly, the electrode material must be applied to the carrier substrate in such a way that this desired low surface roughness is achieved on the electrode. This is opposed by the relatively large surface roughness of the metallic porous carrier substrate. Achieving the desired low surface roughness on the electrode is made even more difficult if the electrode (in particular as an anode) is produced on the carrier substrate by means of sintering under reduced process conditions, because this results in a coarser roughness on the electrode surface. According to the invention, these problems are solved in that the average pore size of the adaptation layer is smaller than the average pore size of the electrode. This ratio of the average pore sizes applies at least to the near-surface layer regions of the surfaces of the electrolyte facing Electrode layer and adaptation layer. This ratio preferably applies to the entire layer thickness of the electrode and the adaptation layer. With the help of the aforementioned ratio of the average pore size, a surface structure is made available in the arrangement which technically simplifies the application of a gas-tight thin-layer electrolyte. In particular, it is possible by means of the adaptation layer to apply particularly thin electrolyte layers (<10 μm) in a gas-tight manner, for example by means of physical vapor deposition (PVD = physical vapor deposition), in particular by means of electron beam evaporation or sputtering processes or sol-gel processes , Depending on the material of the adaptation layer, a single thin electrolyte layer is therefore sufficient for the proper functioning of the fuel cell, which simplifies the manufacture of the fuel cell. In addition, the internal cell resistance of the fuel cell is significantly reduced compared to fuel cells with plasma-sprayed electrolytes, which require a layer thickness of approx. 40 µm for sufficient gas tightness, which means that higher power yields can be achieved.

The adaptation layer can be selected in terms of material and structure, in particular pore structure, such that an electrolyte can always be applied to an electrode (anode or cathode) with the adaptation layer interposed.

According to the invention, the adaptation layer is used in the case of reduced anode layer structures on which a gas-tight electrolyte layer cannot be applied directly. Such reduced anode layer structures arise, for example, in connection with metallic substrates. These substrates are preferably produced by powder metallurgy and in particular are provided in the form of plates. A central area of this substrate is porous and serves as a mechanical support for the electrochemically active cell layers. These cell layers can be, for example, by wet chemical coating (such as screen printing or wet powder spraying) with subsequent sintering or by thermal spraying processes (such as plasma spraying or high-speed flame spraying) getting produced. Metallic carrier substrates have the advantage over ceramic carrier substrates that they are thermally more resilient and more redox-stable during operation. However, oxidation of the carrier substrate during production must be prevented since the formation of metal oxide would cause volume changes in the carrier substrate, which would endanger the defect-free application of the electrodes and the electrolyte to the carrier substrate. In addition, the electrical resistance of the oxidizing carrier substrate increases, which would have an adverse effect on the later cell performance. For this reason, the anode structure applied to the carrier substrate is sintered in a reduced atmosphere, so that the anode structure is in a reduced, porous form. The nickel oxide contained in the anode structure before sintering is reduced during the sintering, which leads to a coarsening of its grain size due to the high sintering activity, and pores with relatively large diameters (for example 2 μm) are formed. Such a surface structure of the anode is often not suitable for applying a gas-tight thin-film electrolyte directly to the anode structure. In particular, the desired gas tightness of the electrolyte is not guaranteed if it is to be applied to the anode structure by means of gas phase deposition (eg PVD process). This problem is solved by means of the adaptation layer described above.

The roughness can be used to physically characterize a surface. The primary profile was measured optically (confocal laser topograph) and the filtered roughness profile and the roughness values were calculated in accordance with DIN EN ISO 11562 and 4287. The lengths of the sensing distance ( I _t ), measuring distances ( I _n ) and individual measuring distances ( I _r ) were selected in accordance with DIN EN ISO 4288. According to DIN EN ISO 4287 _, the arithmetic mean roughness value R _a indicates the arithmetic mean of the amounts of all profile values of a roughness profile. The quadratic mean roughness value R _q (also referred to as mean surface roughness R _q ) is the quadratic mean value of all profile values and weights outliers more than the arithmetic mean roughness value R _a . The average roughness depth R _z is defined according to DIN EN ISO 4287 as the arithmetic mean of the individual roughness depths of all individual measuring sections. A Single roughness means the distance between the highest peak and the deepest groove in a single measurement section. The entire measuring section is divided into 5 consecutive segments of the same size (individual measuring sections). Since the R _z value is determined by the deepest valleys and highest peaks, this is particularly dependent on the measurement method used. In contrast to the optical method used here, it must be taken into account, for example, with mechanical stylus methods that depending on the tip geometry used, not all of the pointed valleys can be captured.

${\mathrm{R}}_{\mathrm{q}}^{\mathrm{\mu }}$

und ${\mathrm{R}}_{\mathrm{z}}^{\mathrm{\mu }}$

gekennzeichnet.In DIN EN ISO 4288 the division of the primary profile into a ripple component (long wavelengths) neglected for the roughness value calculation and into the actual roughness component (small wavelengths) is determined by means of a filter limit wavelength depending on the roughness values achieved. For example, a limit wavelength λ _c of 0.8 mm is provided for an arithmetic mean roughness R _a greater than 0.02 µm and less than or equal to 2.00 µm (with I _r = λ _c ). In particular for layers deposited from the gas phase (PVD), however, bumps of this wavelength do not yet play a decisive role in the quality and tightness of the layer, but bumps with a significantly smaller wavelength. For this reason, in addition to the roughness according to DIN, a so-called micro-roughness is used in this invention, which is based on a limit wavelength of 0.15 mm for otherwise identical total measuring sections. The number of individual measuring sections (normally 5) increases accordingly, since I _r = λ _c must always apply. These micro roughnesses were measured accordingly ${\mathrm{R}}_{\mathrm{a}}^{\mathrm{\mu }}.$

${\mathrm{R}}_{\mathrm{q}}^{\mathrm{\mu }}$

and ${\mathrm{R}}_{\mathrm{z}}^{\mathrm{\mu }}$

characterized.

The average pore size and the sintered grain size can be used as further characteristic parameters for describing the properties of a sintered layer. Both dimensions can be determined for any structure, including open pores, using the line-cutting method on scanning electron microscope images of cross sections. For this purpose, the individual phases (particles, pores) are first of all shown in the images via contrast differences, grain shape or element analysis (e.g. energy dispersive X-ray spectroscopy, EDX) marked accordingly, then statistically drawn straight lines and the intersections at the transitions between the different phases marked. The average value of all lengths of the route sections thus created, which lie in one phase, reflects the average cutting line length for this phase (e.g. pores). This average cutting line length is converted into the actual grain size or pore size by multiplication with a corresponding geometry factor. The value 1.68 and the value 1.56 [2] were used as the geometry factor, assuming the commonly used model of pores around tetrakaidecahedral grains according to reference [1].
If sintered grain sizes are also spoken of in this invention, this means the morphologically readable grain size from the structure. The samples were not etched before analysis.

The maximum pore size was determined from a series of scanning electron micrographs from the largest inner diameters of all pores. The inner diameter of a pore denotes the length of the largest straight section that runs inside the pore.
For the pore and grain size to be determined, a suitable magnification must be observed in the microscopic images. In particular, the pore or grain size to be determined still has to be resolved and at the same time completely captured by the image section.

As already mentioned, the adaptation layer allows the electrolyte to be applied directly, so that in the sense of a simplified, space-saving construction of the fuel cell, additional intermediate layers between the electrolyte and the adaptation layer can be dispensed with.

The average pore size of the adaptation layer is preferably at most half the size of the average pore size of the electrode. It is also possible to use a gas-tight thin-layer electrolyte (<10 µm) via PVD, here to be applied in particular by means of electron beam evaporation or sputter processes, or sol-gel technologies.

The average pore size of the pores (at least in the layer region close to the layer surface facing the electrolyte) of the adaptation layer is preferably at most 500 nm. This supports a homogeneous growth of the electrolyte material (e.g. as a PVD layer) on the adaptation layer. With medium pore sizes above 500 nm there is a risk that the pores can no longer be sealed gas-tight with a thin electrolyte layer. In particular, the average pore size of the adaptation layer (at least in its near-surface layer area of the layer surface facing the electrolyte) is at most 350 nm, more preferably at most 250 nm.

The adaptation layer preferably has, as the average surface roughness, a square mean roughness value R _{q of} less than 2.5 μm, preferably at most 1.5 μm, more preferably at most 1.0 μm. A quadratic mean roughness value R _q above 2.5 µm leads to potential leaks in the subsequent thin-film electrolyte. For example, intercolumnar gaps can arise when a subsequent PVD layer grows. In the case of sol-gel thin-film electrolytes, higher roughness values mean that the wetting of the profile tips can no longer be guaranteed or the critical layer thickness in the profile valleys is exceeded, which leads to cracks in the thin-film electrolyte.

A diffusion barrier is preferably arranged between the carrier substrate and an electrode, in particular the anode. It can prevent metallic interdiffusion and other reactions between the substrate and the electrode and thus contributes to long-term stability and a longer service life of the arrangement.

The adaptation layer preferably has a thickness of 3 to 20 μm. Below a layer thickness of 3 µm, the adaptation layer cannot fully compensate for the roughness of the underlying electrode layer, which is a gas-tight application of a thin-film electrolyte with homogeneous layer growth is not possible. Above a layer thickness of 20 µm, the ohmic resistance of this layer system (adaptation layer and electrolyte) would be in a range that would not offer any significant performance advantage over conventional metal-based SOFCs (= Solid Oxide Fuel Cell) with plasma-sprayed electrolytes.

The electrolyte applied to the adaptation layer preferably has a layer thickness of 0.2 to 10 μm. The required gas tightness of the electrolyte layer is not guaranteed below a layer thickness of 0.2 µm. The increase in the layer thickness of the electrolyte is accompanied by a significant increase in the ohmic resistance and consequently by a reduced output of the fuel cell, so that a maximum layer thickness of 10 μm is preferred.

The arrangement with the electrolyte and the adaptation layer is preferably used in a fuel cell, in particular in a high-temperature fuel cell. High-temperature fuel cells include oxide-ceramic fuel cells - also called SOFC. Due to its high electrical efficiency and the possible use of the waste heat generated at high operating temperatures, the SOFC is particularly suitable as a fuel cell.

A suitable material for the metallic substrate is, for example, a ferritic FeCrMx alloy and a chromium-based alloy. In addition to iron, the FeCrMx alloy regularly has chromium contents between 16 and 30% by weight and additionally at least one alloying element in a proportion of 0.01 to 2% by weight, which is selected from the group of rare earth metals or their oxides, e.g. B. Y, Y ₂ O ₃ , Sc, Sc ₂ O ₃ , or from the group Ti, Al, Mn, Mo or Co.

Examples of suitable ferritic steels are Ferrochrom (1.4742), CrAl20 5 (1.4767) and CroFer 22 APU from Thyssen Krupp, FeCrAlY from Technetics, ZMG 232 from Hitachi Metals, SUS 430 HA and SUS 430 Na from Nippon Steel and all ODS iron-based alloys of the ITM class from Plansee, such as ITM Fe-26Cr- (Mo, Ti, Y ₂ O ₃ )

Alternatively, a chromium-based alloy, that is to say with a chromium content of more than 65% by weight, for example Cr5FeIY or Cr5FeIY ₂ O ₃ , can also be used as the porous metallic substrate.

1. 1) optional eine Diffusionsbarriereschicht (zum Verhindern von metallischer Interdiffusion zwischen Substrat und Elektrode, insbesondere bei Anoden),
2. 2) eine erste Elektrode (Anode oder Kathode),
3. 3) ein Elektrolyt,
4. 4) optional eine Diffusionsbarriereschicht zur Verhinderung von Reaktionen zwischen Elektrolyt und Elektrode, insbesondere bei Hochleistungskathoden aus LSCF (Lanthan-Strontium-Cobalt-Ferrit),
5. 5) eine zweite Elektrode (Kathode oder Anode).

Individual layers of the fuel cell are applied to the provided metallic porous substrate. The following functions or layers are preferably applied in succession:

1. 1) optionally a diffusion barrier layer (to prevent metallic interdiffusion between substrate and electrode, especially in the case of anodes),
2. 2) a first electrode (anode or cathode),
3. 3) an electrolyte,
4. 4) optionally a diffusion barrier layer to prevent reactions between the electrolyte and the electrode, particularly in the case of high-performance cathodes made of LSCF (lanthanum strontium cobalt ferrite),
5. 5) a second electrode (cathode or anode).

The diffusion barrier layer consists, for example, of lanthanum strontium manganite (LSM), lanthanum strontium chromite (LSCR) or gadolinium oxide-doped cerium oxide (CGO). The anode can be constructed as a multilayer composite or as a single layer. The same applies in principle to the cathode. First a first electrode is applied to the substrate, e.g. by means of a wet chemical process.

As already explained, a porous adaptation layer is applied to the electrode. The electrolyte can be applied to the adaptation layer in a gas-tight manner with little outlay on the process, since the average pore size of the adaptation layer is smaller than the average pore size of the electrode.

A suitable layer thickness of the adaptation layer is advantageously achieved by applying it to the electrode by wet chemistry. This can for example by means of screen printing, dip coating or slip casting.

Optionally, the adaptation layer can also be applied in multiple layers. In this case, the material of the adaptation layer is repeatedly applied in several process steps. For example, the electrode is repeatedly dip-coated and dried between individual coating processes. The multi-layer application supports a homogeneous adaptation layer. Irregular surface courses of the adaptation layer are avoided. This in turn creates advantageous physical conditions for the application of the electrolyte material to the adaptation layer.

In a preferred embodiment, the adaptation layer consists of a pure ion-conducting material, that is, an electron-non-conducting material. The required electrical insulation between the two electrodes (anode and cathode) is thus already ensured by the adaptation layer. Additional electronic insulation layers can be omitted, so that the structure of the fuel cell is simplified. The gas-tight electrolyte can therefore also consist of a layer that - e.g. under operating conditions of the fuel cell - has a significant electronic conductivity. This is e.g. for an electrolyte made of gadolinium oxide-doped cerium oxide (CGO) at higher temperatures (> 650 ° C).

Doped zirconium oxide is used as the material for the electronically non-conductive adaptation layer. At least one oxide of the doping elements from the group Y, Sc, Al, Sr, Ca is suitable as the doping. The adaptation layer can be designed as a YSZ layer (yttrium oxide-stabilized zirconium dioxide).

Alternatively, an ion and electron conducting material (mixed conductor) is used for the adaptation layer. Doped cerium oxide is particularly suitable for this. At least one oxide of the doping elements from the group of rare earth elements such as Gd, Sm and / or is advantageous as the doping provided from the group Y, Sc, AI, Sr, Ca. The adaptation layer can be designed as a CGO layer. In this case, the electrical insulation between the two electrodes should be taken over by the gas-tight electrolyte layer. An oxide ceramic, for example doped zirconium oxide, is preferably used as the material for the electronically non-conductive thin-film electrolyte. At least one oxide of the doping elements from the group Y, Sc, Al, Sr, Ca is suitable as doping. The thin-layer electrolyte can be designed as a YSZ layer (yttrium oxide-stabilized zirconium dioxide).
Depending on the application, the aforementioned materials for the adaptation layer can also be used for the electrolyte. In the case of an electrolyte made of CGO material, cathodes can also be applied directly to this electrolyte, which are designed as a Sr component reacting with ZrO ₂ , for example lanthanum strontium cobalt ferrite (LSCF) or lanthanum strontium cobaltite ( LSC).

The adaptation layer applied to the electrode is preferably sintered. The sintering temperature is in particular 950 ° C to 1300 ° C, so that no undesired structural changes in the adaptation layer can be expected during operation of the fuel cell (e.g. SOFC, up to 850 ° C). In order to achieve sufficient mechanical stability, a powder with an average grain size of 30 to 500 nm, in particular 150 nm, is preferably used for the adaptation layer. This also prevents excessive infiltration into a porous electrode layer (e.g. anode layer).

The adaptation layer offers the possibility of producing a stable and gas-tight electrolyte layer structure by means of gas phase deposition. This process also allows particularly thin electrolyte layers. For example, an electrolyte with a layer thickness of 0.2 to 10 μm, preferably 1 to 3 μm, more preferably 1 to 2 μm, can be deposited on the adaptation layer. The PVD process (physical vapor deposition) is particularly suitable for this.

Alternatively, the electrolyte can be applied using sol-gel technology.

The invention is explained in more detail below with the aid of a few figures and a specific exemplary embodiment.

The Figure 1 shows the surface of a reduced anode structure (Ni / 8YSZ), which is applied to a porous metallic substrate (ITM), not shown here. The geometry / dimensioning of the pores on the anode structure is relatively large.

Figure 2 shows a cross section of the anode structure coated with an electrolyte Figure 1 , The multilayer electrolyte was applied to the anode structure by means of a PVD coating and consists of a CGO layer (E1), an 8YSZ layer (E2) and a further CGO layer (E3). The columnar layer growth of the electrolyte with a fanned, irregular growth is clearly recognizable. The inhomogeneous growth of the electrolyte layers, in particular on the Ni particles, prevents the required gas-tight layer structure of the electrolyte.

The Figure 3 shows the surface structure of the adaptation layer applied to an anode structure. This is clearly recognizable in comparison with the anode structure Figure 1 significantly reduced pore size of the pores of the adaptation layer.
The Figure 4 shows a cross section of the adaptation layer according to Figure 3 and an electrolyte applied thereon. The electrolyte is formed as a single layer of CGO and was applied using a PVD process. The growth of the electrolyte layer is undisturbed and homogeneous, so that the required gas tightness of the electrolyte is achieved.

Examples of the structure of the arrangement or the fuel cell according to the invention are the Figures 5 and 6 schematically removable.

S:

FeCr-Legierung oder CFY-Legierung.

D:

Diffusionsbarriere aus LSM oder CGO.

A:

Ni/8YSZ (Cermetgemisch aus Nickel und einem mit 8 Mol-% Yttriumoxid stabilisiertem Zirkondioxid) oder NiO/8YSZ (Gemisch aus Nickeloxid und einem mit 8 Mol-% Yttriumoxid stabilisiertem Zirkondioxid).

AD:

YSZ (Yttriumoxid-stabilisiertes Zirkoniumdioxid) oder ScSZ (Scandiumoxid-stabilisiertes Zirkoniumdioxid).

E:

CGO.

K:

LSCF oder LSM oder LSC.

According to Fig. 5 (Variant A) is a porous anode structure on a metallic porous substrate S (ITM) which is provided with a diffusion barrier D. A applied. The following layers are applied in succession to this anode structure: a porous adaptation layer AD, a gas-tight electrolyte layer E, a porous cathode K. The following materials are used in this construction, for example:

S:

FeCr alloy or CFY alloy.

D:

Diffusion barrier made of LSM or CGO.

A:

Ni / 8YSZ (cermet mixture of nickel and a zirconium dioxide stabilized with 8 mol% yttrium oxide) or NiO / 8YSZ (mixture of nickel oxide and a zirconium dioxide stabilized with 8 mol% yttrium oxide).

AD:

YSZ (yttria-stabilized zirconia) or ScSZ (scandia-stabilized zirconia).

e:

CGO.

K:

LSCF or LSM or LSC.

S:

FeCr-Legierung oder CFY-Legierung.

K:

LSM oder LSCF oder LSC.

AD:

CGO.

E:

YSZ oder ScSZ.

A:

Ni/8YSZ oder NiO/8YSZ.

According to Fig. 6 (Variant B), a porous cathode K is applied to a metallic porous substrate S (ITM). The following layers are applied in succession to this cathode K: a porous adaptation layer AD, a gas-tight electrolyte layer E, a porous anode A. The following materials are used in this construction, for example:

S:

FeCr alloy or CFY alloy.

K:

LSM or LSCF or LSC.

AD:

CGO.

e:

YSZ or ScSZ.

A:

Ni / 8YSZ or NiO / 8YSZ.

kleiner als 1 µm, vorzugsweise kleiner als 0,6 µm, und für die mittlere Rautiefe R_z kleiner 10 µm, vorzugsweise kleiner 6 µm, sowie für die mittlere Mikrorautiefe ${\mathrm{R}}_{\mathrm{z}}^{\mathrm{\mu }}$

kleiner 4 µm, vorzugsweise kleiner 2 µm betragen.The application of a gas-tight thin-layer electrolyte places certain demands on the underlying layer structure with regard to roughness and / or pore size, which can be fulfilled by an adaptation layer. If powder-metallurgical porous substrates (eg with a grain size of <125 µm) are coated with an anode structure, the latter can have a medium one Pore size up to 1.5 µm (see Figure 1 ). The roughness of the surface of this anode structure should be less than 3 μm, preferably less than 2 μm, for the square micro-center roughness value for the square mean roughness value R _q ${\mathrm{R}}_{\mathrm{q}}^{\mathrm{\mu }}$

less than 1 µm, preferably less than 0.6 µm, and for the mean roughness depth R _z less than 10 µm, preferably less than 6 µm, and for the mean microroughness depth ${\mathrm{R}}_{\mathrm{z}}^{\mathrm{\mu }}$

less than 4 microns, preferably less than 2 microns.

The laser topograph CT200 (Cybertechnolgies GmbH, Ingolstadt) with a confocal laser sensor LT9010 was used to determine the roughness (measuring spot size approx. 2µm, vertical resolution 10nm). The primary profiles measured in 1 µm increments were filtered with a Gaussian filter, α = ln (2), filter length 5 µm, in order to minimize individual false signals due to multiple reflections before the DIN regulations were applied.

At least three scanning electron micrographs of cross sections of the layers were evaluated for each parameter for the grain and pore sizes of the sintered structure determined using the line cutting method. 500-1,000 lines were drawn per image. With a number of pixels of the raster electronic recordings of 1024x768 pixels, a total section of the width of 5 to 15 µm was chosen for the adaptation layer.

zeigte einen Wert von 0,21 µm und die gemittelte Mikrorautiefe ${\mathrm{R}}_{\mathrm{z}}^{\mathrm{\mu }}$

zeigte einen Wert von 0,67 µm. Neben dieser leichten Absenkung der Rauhigkeitswerte zeigte sich eine deutliche Verringerung der mittleren Porengröße an der Oberfläche der Adaptionsschicht. Während die Oberfläche der Anodenstruktur eine mittlere Porengröße von ca. 610 nm, aufwies (siehe Figur 1), betrug die mittlere Porengröße der Adaptionsschicht in diesem Fall ca. 240 nm (siehe Figur 3). Auf der Adaptionsschicht konnte eine dichte Elektrolyt-Schicht aus Gd₂O₃-dotiertem CeO₂ (CGO) mittels Gasphasenabscheidung (Elektronenstrahlverdampfung bei 870°C, EB-PVD) mit einer Schichtdicke von ca. 1,7 µm aufgebracht werden. Die Gasdichtheit dieses Elektrolyt wurde mittels He-Lecktest zu 3,4 x 10^-3 (hPa dm³) / (s cm²) für eine Druckdifferenz von 1000 hPa ermittelt. Dieser Wert entspricht gängigen anodengestützten Brennstoffzellen im reduzierten Zustand.An 8YSZ powder with an average dispersible primary particle size of 150 nm and a specific surface area of 13 m ² / g was used for the adaptation layer (TZ-8Y, Tosoh Corp., Japan). An immersion suspension was mixed with grinding balls with a diameter of 5 and 10 mm and homogenized on a roller bench for 48 hours, consisting of 67.2% by weight of solvent DBE (Dibasic ester, Lemro Chemical Products Michael Mrozyk KG, Grevenbroich), 30.5% by weight. % 8YSZ powder (TZ-8Y) and 2.3% by weight of ethyl cellulose as a binder (Fluka, 3-5.5 mPa s, Sigma-Aldrich Chemie GmbH, Munich). The carrier substrates with the anode structure applied to them were immersed vertically in the suspension and, after a drying step in an H ₂ atmosphere, sintered at 1200 ° C. for 3 hours. Depending on the coating parameters (Immersion speed, dripping time), an adaptation layer thickness of 10 to 20 µm was obtained. The adaptation layer applied in this way had a square mean roughness value R _q of 1.2 μm and an average roughness depth R _z of 5.8 μm. The square micro-center roughness ${\mathrm{R}}_{\mathrm{q}}^{\mathrm{\mu }}$

showed a value of 0.21 µm and the average microroughness ${\mathrm{R}}_{\mathrm{z}}^{\mathrm{\mu }}$

showed a value of 0.67 µm. In addition to this slight reduction in the roughness values, there was a significant reduction in the average pore size on the surface of the adaptation layer. While the surface of the anode structure had an average pore size of approx. 610 nm (see Figure 1 ), the average pore size of the adaptation layer in this case was approx. 240 nm (see Figure 3 ). A dense electrolyte layer of Gd ₂ O ₃ -doped CeO ₂ (CGO) could be applied to the adaptation layer by means of gas phase deposition (electron beam evaporation at 870 ° C, EB-PVD) with a layer thickness of approx. 1.7 µm. The gas tightness of this electrolyte was determined by means of a He leak test to be 3.4 × 10 ^-3 (hPa dm ³ ) / (s cm ² ) for a pressure difference of 1000 hPa. This value corresponds to common anode-supported fuel cells in the reduced state.

Literature cited in the application:

• [1] TS Smith: "Morphological Characterization of Porous Coatings." In: "Quantitative Characterization and Performance of Porous Implants for Hard Tissue Applications", ASTM STP953, JE Lemmons, ed., American Society for Testing and Materials, Philadelphia, 1987, pp. 92-102 ,
• [2] MI Mendelson: "Average Grain Size in Polycrystalline Ceramics", J. Am. Ceram. Soc. 52 [8] (1969), 443-446 ,

Claims (13)

1. An arrangement for a fuel cell, having

   - at least one electrode (A, K), present in reduced form

   - an electrolyte (E),

   - a metallic porous carrier substrate (8) as a carrier for the electrode (A, K) and the electrolyte,

   - an adaptation layer (AD) arranged between the reduced electrode present and the electrolyte for adapting the electrolyte (E) to this electrode (A, K),
   characterised in that

   - the adaptation layer (AD) has an average pore size which is smaller than the average pore size of the reduced electrode present (A, K),
   and

   - having an adaptation layer (AD) made from doped zirconium oxide, wherein the doping contains at least one oxide of the doping elements from the group Y, Sc, Al, Sr, Ca, or

   - having an adaptation layer (AD) made from doped cerium oxide, wherein the doping contains at least one oxide of the doping elements from the group of the rare earth elements and/or from the group Y, Sc, Al, Sr, Ca

   - wherein the electrolyte is a gastight thin layer electrolyte with a maximum layer thickness of 10 µm.
2. The arrangement according to claim 1, characterised in that the average pore size of the adaptation layer (AD) is at most half as large as the average pore size of the electrode (A, K).
3. The arrangement according to claim 1 or 2, characterised in that the average pore size of the adaptation layer (AD) is at most 500 nm, preferably at most 350 nm.
4. The arrangement according to one of the preceding claims, characterised in that the adaptation layer (AD) has an average surface roughness R_q smaller than 2.5 µm, preferably at most 1.5 µm, more preferably at most 1.0 µm.
5. The arrangement according to one of the preceding claims, characterised in that the electrode is designed as an anode (A).
6. The arrangement according to one of the preceding claims, characterised in that the electrolyte (E) is directly arranged on the layer surface of the adaptation layer (AD), the layer surface facing towards the electrolyte (E).
7. The arrangement according to one of the preceding claims, characterised in that the adaptation layer (AD) has a thickness from 3 to 20 µm, preferably from 3 to 7 µm.
8. The arrangement according to one of the preceding claims, characterised in that the electrolyte (E) has a thickness from 0.2 to 10 µm, preferably from 1 to 3 µm.
9. The arrangement according to one of the preceding claims, characterised by an electrolyte (E) made from a material that is non-conductive for electrons.
10. The arrangement according to claim 9, characterised by an electrolyte (E) made from doped zirconium oxide, wherein the doping contains at least one oxide of the doping elements from the group Y, Sc, Al, Sr, Ca.
11. The arrangement according to one of the preceding claims 1 to 8, characterised by an electrolyte (E) made from a material that is conductive for ions and electrons.
12. The arrangement according to claim 11, characterised by an electrolyte (E) made from doped cerium oxide, wherein the doping contains at least one oxide of the doping elements from the group of the rare earth elements such as Gd, Sm, and/or from the group Y, Sc, Al, Sr, Ca.
13. A method for producing an arrangement for a fuel cell according to one of claims 1-12, having an electrode (A, K) and an electrolyte (E), characterised by the following method steps:

    a) provision of a metallic, porous carrier substrate (8) as a carrier for the electrode (A, K) and the electrolyte (E),

    b) application of the electrode (A, K) to the carrier substrate (8),

    c) sintering of the electrode structure (A, K) applied to the carrier substrate (8) in a reduced atmosphere,

    d) application of a porous adaptation layer (AD) to the electrode (A, K) for adapting the electrolyte (E) to this electrode (A, K), wherein the average pore size of the adaptation layer (AD) is smaller than the average pore size of this electrode (A, K), and

    e) application of the electrolyte (E) to the adaptation layer (AD).

Images